INDEX

Explanations

references to web or app content and updates

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 35-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_35/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.35.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

XtraReports

-0.52

 nakalista

-0.52

хьтан

-0.51

Hentet

-0.48

xase

-0.47

tonsoft

-0.46

}-*/;

-0.46

timewa

-0.45

 Normdatei

-0.45

DIPSETTING

-0.44

POSITIVE LOGITS

 style

0.41

 scolaires

0.41

 kait

0.38

melidir

0.38

 estri

0.38

 escolares

0.37

 yön

0.37

 lemah

0.37

 alambre

0.37

 Escolar

0.36

Activations Density 0.242%

references to web or app content and updates

No Comments

No Known Activations